CtrlK
BlogDocsLog inGet started
Tessl Logo
PluginPublic

ainativedev/aidevcon-2026-nyc

AI Native DevCon 2026 New York — practical agent skills grounded in conference talks

Quality

86%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Low

Low-risk findings worth noting

Overview
Quality
Evals
Security
Files

outline.mdskills/talk-katsioloudes-montei-agentic-lifecycle/references/

Source outline

Source: Joseph Katsioloudes & Peyton Montei, Rebuilding the Software Lifecycle for Coding Agents and Humans, AI Native DevCon New York 2026. Recording. All anchors refer to original numbered cues in the transcript. Short quotations below preserve the supplied transcription, including spoken wording.

TopicSource anchorExact short quoteApplication and scope
Review burden claims00:00:55–00:02:36, cues 13–36“This data comes from LinearB.”Industry figures are cited by the speakers, not independently verified here.
Intent and decisions00:02:43–00:04:14, cues 37–56“what was intended, what was decided, and what was verified.”A diff lacks important process context.
Trail overview00:05:57–00:08:57, cues 79–116“on each commit”The demo associates evaluations with change history.
Review loop00:09:03–00:09:48, cues 117–125“loop until all findings are resolved.”The demonstrated request uses an iterative review loop.
Custom judgments00:09:59–00:12:06, cues 127–155“How much? My intent is matched”Review criteria can include intent and personal risk preferences.
Nondeterminism and gates00:14:53–00:15:54, cues 191–205“very non-deterministic”Prompt judgments differ from deterministic checks and gates.
Project-level grouping00:16:45–00:18:06, cues 207–223“multiple trails.”The demo groups related changes and sessions.
Memory provenance00:18:12–00:19:54, cues 224–245“the link to the trail that created those”Memories retain a connection to originating work.
Memory review00:20:39–00:21:36, cues 254–268“reviewing memories at the same time as you review code”Durable guidance should be reviewed alongside changes.
Session evidence00:23:34–00:24:29, cues 294–307“Here are the tool calls”The demo exposes session and tool history.
Conflict demonstration00:25:30–00:28:21, cues 320–359“If you want me to honor the stored no red button preference,”The live example surfaces a preference conflict without proving a general policy.

Source limitations

  • Entire features and the live flow are talk-time demonstrations, not current product guarantees or a complete integration guide.
  • LinearB statistics and distribution-speed claims are reported comparisons with insufficient methodology in this source; do not infer universal productivity or latency gains.
  • Prompt-based evaluations are explicitly nondeterministic. A score or resolved-finding status alone does not establish correctness or release authority.
  • The conflict demo's observed wording is nuanced; do not claim it deterministically rejected the red button or implemented a universal precedence rule.
  • This is a spoken transcript; unseen slides, linked repositories, and third-party research were not independently inspected. No outcome beyond the described talk is established.

skills

talk-katsioloudes-montei-agentic-lifecycle

tile.json